Yojana Magazine June 2022 - 2 | Monthly Yojana & Kurukshetra Magazine (English) - UPSC PDF Download

 Page 1


YOJANA   June 2022 17
ndia has a vibrant startup ecosystem with 
supporting infrastructure– incubators, 
development grants, angel/venture investors, 
mentors– and a conducive policy environment. 
The Economic Survey of India 2021-22 says that there 
are 61,400 registered startups in India, making it the  
third-largest startup ecosystem in the world behind China 
and US. Around 14,000
1
 new startups were registered 
in India during CY2021
2
. Over the past decade, Indian 
startups have created 6.6 lakh direct jobs and 34 lakh 
indirect jobs. 
Indian startups raised USD 24 billion in CY21, 
compared to USD 10 billion in CY20. There has been a 
significant localisation and diversification in the investor 
pool for startups in India over the past decade. There were 
more than 750 institutional investors in India in CY21, 
80% more than in CY20. The number of angel investors 
grew in CY21 by 20% to about 2,400. More than half the 
investment deals in CY21 had an India-based investor. 
Over 250 corporates have engaged with Indian startups 
in some way, including by running 80+ open innovation 
programmes for startups in CY21.
The Central and State governments in India have 
actively supported the startup sector over the past decade. 
The Startup India platform, which started in 2016, has 
been instrumental in encouraging startups and integrating 
them with the corporate and investment community. Over 
26 States in India have a startup policy.
What is a Deep-Tech Startup? 
Notwithstanding the healthy development of India’s 
startup ecosystem, one weakness that keeps India behind 
the developed countries is that we lack deep-tech startups. 
“Deep-tech” startups constitute less than one per cent of 
innovation Deep-Tech Startup Ecosystem
R Raghuttama Rao
The author is the CEO, Gopalakrishnan-Deshpande Centre for Innovation & Entrepreneurship, IIT Madras. Email: enquiries@gdciitm.org
I
the number of startups, far below what a fast-growing, 
complex, and large economy like India should have. 
The absence of deep-tech startups harms India 
considerably by weakening her capability to meaningfully 
address complex socio-economic challenges that afflict our 
society in multiple sectors such as agriculture, healthcare, 
transportation, education, energy, etc. The solutions to such 
challenges that address the UN’s Sustainable Development 
Goals would necessarily have to be radically new and 
disrupt existing industries and business processes.
In India’s population of 130 crores, only the top 
25%
3
 (affluent and middle-class) benefit from the fruits of 
technological progress, be it healthcare, consumer goods, 
clean water, safe transportation, education, etc. In contrast, 
the remaining 100 crore people do not get enough or are 
substantially bypassed. This is because most of the hi-tech 
goods and services are designed in the developed world 
for rich people– the average per capita income in OECD 
countries is about USD 40,000, while the average per 
capita income of the bottom 100 crore people in India is 
around USD 1000
3
. They simply cannot afford modern 
innovations with an income of 2.5% of the people for 
whom such innovations are designed. So, how do 100 
crore Indians move towards development?
The answer lies in becoming Atmanirbhar in 
commercialising domestic science and technology to solve 
our challenging problems.  
India’s development challenges are so unique and 
idiosyncratic that innovators from developed countries, 
not familiar with our context or cost structures, will not be 
able to provide solutions. The clarion call from the Prime 
Minister for ‘Atmanirbhar’ is apt here– we have to grow 
our own deep-tech ecosystem. 
Deep-tech startups arise from research-based, disruptive innovations from STEM labs of 
academic/research institutions and solve hard problems and challenges. India lacks deep-tech 
startups. Deep-tech startups constitute less than one per cent of the number of startups, far 
below what a fast-growing, complex, and large economy like India should have. 
Page 2


YOJANA   June 2022 17
ndia has a vibrant startup ecosystem with 
supporting infrastructure– incubators, 
development grants, angel/venture investors, 
mentors– and a conducive policy environment. 
The Economic Survey of India 2021-22 says that there 
are 61,400 registered startups in India, making it the  
third-largest startup ecosystem in the world behind China 
and US. Around 14,000
1
 new startups were registered 
in India during CY2021
2
. Over the past decade, Indian 
startups have created 6.6 lakh direct jobs and 34 lakh 
indirect jobs. 
Indian startups raised USD 24 billion in CY21, 
compared to USD 10 billion in CY20. There has been a 
significant localisation and diversification in the investor 
pool for startups in India over the past decade. There were 
more than 750 institutional investors in India in CY21, 
80% more than in CY20. The number of angel investors 
grew in CY21 by 20% to about 2,400. More than half the 
investment deals in CY21 had an India-based investor. 
Over 250 corporates have engaged with Indian startups 
in some way, including by running 80+ open innovation 
programmes for startups in CY21.
The Central and State governments in India have 
actively supported the startup sector over the past decade. 
The Startup India platform, which started in 2016, has 
been instrumental in encouraging startups and integrating 
them with the corporate and investment community. Over 
26 States in India have a startup policy.
What is a Deep-Tech Startup? 
Notwithstanding the healthy development of India’s 
startup ecosystem, one weakness that keeps India behind 
the developed countries is that we lack deep-tech startups. 
“Deep-tech” startups constitute less than one per cent of 
innovation Deep-Tech Startup Ecosystem
R Raghuttama Rao
The author is the CEO, Gopalakrishnan-Deshpande Centre for Innovation & Entrepreneurship, IIT Madras. Email: enquiries@gdciitm.org
I
the number of startups, far below what a fast-growing, 
complex, and large economy like India should have. 
The absence of deep-tech startups harms India 
considerably by weakening her capability to meaningfully 
address complex socio-economic challenges that afflict our 
society in multiple sectors such as agriculture, healthcare, 
transportation, education, energy, etc. The solutions to such 
challenges that address the UN’s Sustainable Development 
Goals would necessarily have to be radically new and 
disrupt existing industries and business processes.
In India’s population of 130 crores, only the top 
25%
3
 (affluent and middle-class) benefit from the fruits of 
technological progress, be it healthcare, consumer goods, 
clean water, safe transportation, education, etc. In contrast, 
the remaining 100 crore people do not get enough or are 
substantially bypassed. This is because most of the hi-tech 
goods and services are designed in the developed world 
for rich people– the average per capita income in OECD 
countries is about USD 40,000, while the average per 
capita income of the bottom 100 crore people in India is 
around USD 1000
3
. They simply cannot afford modern 
innovations with an income of 2.5% of the people for 
whom such innovations are designed. So, how do 100 
crore Indians move towards development?
The answer lies in becoming Atmanirbhar in 
commercialising domestic science and technology to solve 
our challenging problems.  
India’s development challenges are so unique and 
idiosyncratic that innovators from developed countries, 
not familiar with our context or cost structures, will not be 
able to provide solutions. The clarion call from the Prime 
Minister for ‘Atmanirbhar’ is apt here– we have to grow 
our own deep-tech ecosystem. 
Deep-tech startups arise from research-based, disruptive innovations from STEM labs of 
academic/research institutions and solve hard problems and challenges. India lacks deep-tech 
startups. Deep-tech startups constitute less than one per cent of the number of startups, far 
below what a fast-growing, complex, and large economy like India should have. 
18 YOJANA   June 2022
Need for Deep-Tech Startup Ecosystem
The phrase ‘deep-tech startup’ does not have a precise 
definition, but there is a broad consensus on what it is. 
Deep-tech startups arise from research-based, disruptive 
innovations from STEM labs of academic/research 
institutions and solve hard problems and challenges. 
Some examples are— (a) recycling sewage to get clean 
water at an affordable cost, (b) a low-cost solution at scale 
for curing blindness, (c) affordable solutions for treating 
diseases such as diabetes, dementia, cancer, etc., (d) 
creating an alternative to Lithium-ion batteries, and (e) 
low-cost satellite launching systems.
There are three major problems that deep-tech startups 
have vis-à-vis other startups (including those that are called 
tech-based startups).
1. Deep-tech startups need a longer gestation for 
development than other startups. The latter might 
need from 1-3 years to reach revenue, while deep-tech 
startups need 5-8 years.
2. Deep-tech startups require different types of inputs– 
they require more patient capital, specialised talent, 
and expert knowledge in more than one domain, to 
develop and validate a science-based innovation to the 
point where it is acceptable to commercial investors. 
For example, assume an invention involving creating 
a new substance (say a chemical that removes heavy 
metal from water). It takes time and resources to test 
and validate samples, obtain regulatory approvals, 
and set up a new manufacturing process to produce at 
scale. All these are capital-intensive, time-consuming, 
and have no assurance of success.
3. A deep-tech startup follows a different development 
path than other startups. A deep-tech startup derives 
its IP from the underlying science. The startup has to 
work backwards and find a real-life problem that is 
worth solving using its technology and validate the 
adequacy and nature of the market demand for the 
innovation.
Therefore, deep-tech startups 
take more time, talent, and capital 
to develop, upto when commercial 
investors find them acceptable. The 
risk of failure is high at every stage 
for a deep-tech startup, usually higher 
than in the case of other types of 
startups. But the payoffs of successful  
deep-tech startups are tremendous. 
Think of Microsoft, Google, Apple, 
Intel, Tesla, Moderna, SpaceX, etc. 
They are large corporations today, but 
they started as mere technology bets 
not very long ago. 
India has also created a few deep-tech startups over 
the past decade, whose impact has been overwhelmingly 
positive. It lends credence to the suggestion to step up 
policy and financial support to the deep-tech startup 
ecosystem.
Creating Ecosystem
India has produced about 94 unicorns so far, but barely 
any of them can claim to be a deep-tech startup. We have 
several venture funds in India, but most pursue relatively 
‘lower risk’ investment opportunities that exploit India’s 
growing consumption economy or those making cloned 
products. While India has a problem of inadequate R&D 
expenditure for an economy of her size, there is a sufficient 
amount of high-quality research in India’s top STEM 
colleges to fuel a deep-tech startup revolution. Some key 
reasons why our academic researchers lag in their potential 
to convert research into deep-tech startups are:
1. There is inadequate appreciation amongst 
policymakers and university administrators for the 
need to build capacity amongst academic researchers, 
scientists, and STEM students in India to truly 
understand what entrepreneurship entails and what 
commercialisation of research means. Being formally 
trained in science and technology but not having 
adequate exposure to the real world of business/
commerce, academic researchers conflate invention 
and innovation. There is a big difference between 
making a successful technological breakthrough in 
the lab and building a successful enterprise around 
it. Becoming entrepreneurial cannot be imbibed by 
reading or scholastic programmes but only through 
experiential learning and expert mentoring/coaching.
2. While Government has made good efforts to fund 
innovation in universities through programmes such 
as prototype development, filing for IPR, incubation, 
etc., few academics (<5%) commercialise their 
research by startups. A key point is that even if 
academics aspire to convert their inventions into 
enterprises, they do not have the mental make-up 
(the entrepreneur’s mindset) or the 
knowledge of how to organise what 
they have and collaborate with others to 
get what they do not have/know. Many 
universities have set up incubators 
to help with this, but they are not 
adequately equipped or incentivised to 
commercialise research. Although they 
are not-for-profit entities, incubators 
look for startups that have a good 
chance to be commercially viable. With 
their limited budgets, incubators face 
a tough challenge to nurture startups 
to scale their revenues and become 
Deep-tech startups take more 
time, talent, and capital 
to develop, upto when 
commercial investors find 
them acceptable. The risk of 
failure is high at every stage 
for a deep-tech startup, usually 
higher than in the case of 
other types of startups. But the 
payoffs of successful deep-tech 
startups are tremendous.
Page 3


YOJANA   June 2022 17
ndia has a vibrant startup ecosystem with 
supporting infrastructure– incubators, 
development grants, angel/venture investors, 
mentors– and a conducive policy environment. 
The Economic Survey of India 2021-22 says that there 
are 61,400 registered startups in India, making it the  
third-largest startup ecosystem in the world behind China 
and US. Around 14,000
1
 new startups were registered 
in India during CY2021
2
. Over the past decade, Indian 
startups have created 6.6 lakh direct jobs and 34 lakh 
indirect jobs. 
Indian startups raised USD 24 billion in CY21, 
compared to USD 10 billion in CY20. There has been a 
significant localisation and diversification in the investor 
pool for startups in India over the past decade. There were 
more than 750 institutional investors in India in CY21, 
80% more than in CY20. The number of angel investors 
grew in CY21 by 20% to about 2,400. More than half the 
investment deals in CY21 had an India-based investor. 
Over 250 corporates have engaged with Indian startups 
in some way, including by running 80+ open innovation 
programmes for startups in CY21.
The Central and State governments in India have 
actively supported the startup sector over the past decade. 
The Startup India platform, which started in 2016, has 
been instrumental in encouraging startups and integrating 
them with the corporate and investment community. Over 
26 States in India have a startup policy.
What is a Deep-Tech Startup? 
Notwithstanding the healthy development of India’s 
startup ecosystem, one weakness that keeps India behind 
the developed countries is that we lack deep-tech startups. 
“Deep-tech” startups constitute less than one per cent of 
innovation Deep-Tech Startup Ecosystem
R Raghuttama Rao
The author is the CEO, Gopalakrishnan-Deshpande Centre for Innovation & Entrepreneurship, IIT Madras. Email: enquiries@gdciitm.org
I
the number of startups, far below what a fast-growing, 
complex, and large economy like India should have. 
The absence of deep-tech startups harms India 
considerably by weakening her capability to meaningfully 
address complex socio-economic challenges that afflict our 
society in multiple sectors such as agriculture, healthcare, 
transportation, education, energy, etc. The solutions to such 
challenges that address the UN’s Sustainable Development 
Goals would necessarily have to be radically new and 
disrupt existing industries and business processes.
In India’s population of 130 crores, only the top 
25%
3
 (affluent and middle-class) benefit from the fruits of 
technological progress, be it healthcare, consumer goods, 
clean water, safe transportation, education, etc. In contrast, 
the remaining 100 crore people do not get enough or are 
substantially bypassed. This is because most of the hi-tech 
goods and services are designed in the developed world 
for rich people– the average per capita income in OECD 
countries is about USD 40,000, while the average per 
capita income of the bottom 100 crore people in India is 
around USD 1000
3
. They simply cannot afford modern 
innovations with an income of 2.5% of the people for 
whom such innovations are designed. So, how do 100 
crore Indians move towards development?
The answer lies in becoming Atmanirbhar in 
commercialising domestic science and technology to solve 
our challenging problems.  
India’s development challenges are so unique and 
idiosyncratic that innovators from developed countries, 
not familiar with our context or cost structures, will not be 
able to provide solutions. The clarion call from the Prime 
Minister for ‘Atmanirbhar’ is apt here– we have to grow 
our own deep-tech ecosystem. 
Deep-tech startups arise from research-based, disruptive innovations from STEM labs of 
academic/research institutions and solve hard problems and challenges. India lacks deep-tech 
startups. Deep-tech startups constitute less than one per cent of the number of startups, far 
below what a fast-growing, complex, and large economy like India should have. 
18 YOJANA   June 2022
Need for Deep-Tech Startup Ecosystem
The phrase ‘deep-tech startup’ does not have a precise 
definition, but there is a broad consensus on what it is. 
Deep-tech startups arise from research-based, disruptive 
innovations from STEM labs of academic/research 
institutions and solve hard problems and challenges. 
Some examples are— (a) recycling sewage to get clean 
water at an affordable cost, (b) a low-cost solution at scale 
for curing blindness, (c) affordable solutions for treating 
diseases such as diabetes, dementia, cancer, etc., (d) 
creating an alternative to Lithium-ion batteries, and (e) 
low-cost satellite launching systems.
There are three major problems that deep-tech startups 
have vis-à-vis other startups (including those that are called 
tech-based startups).
1. Deep-tech startups need a longer gestation for 
development than other startups. The latter might 
need from 1-3 years to reach revenue, while deep-tech 
startups need 5-8 years.
2. Deep-tech startups require different types of inputs– 
they require more patient capital, specialised talent, 
and expert knowledge in more than one domain, to 
develop and validate a science-based innovation to the 
point where it is acceptable to commercial investors. 
For example, assume an invention involving creating 
a new substance (say a chemical that removes heavy 
metal from water). It takes time and resources to test 
and validate samples, obtain regulatory approvals, 
and set up a new manufacturing process to produce at 
scale. All these are capital-intensive, time-consuming, 
and have no assurance of success.
3. A deep-tech startup follows a different development 
path than other startups. A deep-tech startup derives 
its IP from the underlying science. The startup has to 
work backwards and find a real-life problem that is 
worth solving using its technology and validate the 
adequacy and nature of the market demand for the 
innovation.
Therefore, deep-tech startups 
take more time, talent, and capital 
to develop, upto when commercial 
investors find them acceptable. The 
risk of failure is high at every stage 
for a deep-tech startup, usually higher 
than in the case of other types of 
startups. But the payoffs of successful  
deep-tech startups are tremendous. 
Think of Microsoft, Google, Apple, 
Intel, Tesla, Moderna, SpaceX, etc. 
They are large corporations today, but 
they started as mere technology bets 
not very long ago. 
India has also created a few deep-tech startups over 
the past decade, whose impact has been overwhelmingly 
positive. It lends credence to the suggestion to step up 
policy and financial support to the deep-tech startup 
ecosystem.
Creating Ecosystem
India has produced about 94 unicorns so far, but barely 
any of them can claim to be a deep-tech startup. We have 
several venture funds in India, but most pursue relatively 
‘lower risk’ investment opportunities that exploit India’s 
growing consumption economy or those making cloned 
products. While India has a problem of inadequate R&D 
expenditure for an economy of her size, there is a sufficient 
amount of high-quality research in India’s top STEM 
colleges to fuel a deep-tech startup revolution. Some key 
reasons why our academic researchers lag in their potential 
to convert research into deep-tech startups are:
1. There is inadequate appreciation amongst 
policymakers and university administrators for the 
need to build capacity amongst academic researchers, 
scientists, and STEM students in India to truly 
understand what entrepreneurship entails and what 
commercialisation of research means. Being formally 
trained in science and technology but not having 
adequate exposure to the real world of business/
commerce, academic researchers conflate invention 
and innovation. There is a big difference between 
making a successful technological breakthrough in 
the lab and building a successful enterprise around 
it. Becoming entrepreneurial cannot be imbibed by 
reading or scholastic programmes but only through 
experiential learning and expert mentoring/coaching.
2. While Government has made good efforts to fund 
innovation in universities through programmes such 
as prototype development, filing for IPR, incubation, 
etc., few academics (<5%) commercialise their 
research by startups. A key point is that even if 
academics aspire to convert their inventions into 
enterprises, they do not have the mental make-up 
(the entrepreneur’s mindset) or the 
knowledge of how to organise what 
they have and collaborate with others to 
get what they do not have/know. Many 
universities have set up incubators 
to help with this, but they are not 
adequately equipped or incentivised to 
commercialise research. Although they 
are not-for-profit entities, incubators 
look for startups that have a good 
chance to be commercially viable. With 
their limited budgets, incubators face 
a tough challenge to nurture startups 
to scale their revenues and become 
Deep-tech startups take more 
time, talent, and capital 
to develop, upto when 
commercial investors find 
them acceptable. The risk of 
failure is high at every stage 
for a deep-tech startup, usually 
higher than in the case of 
other types of startups. But the 
payoffs of successful deep-tech 
startups are tremendous.
YOJANA   June 2022 19
There is a big difference 
between making a successful 
technological breakthrough 
in the lab and building a 
successful enterprise around 
it. Becoming entrepreneurial 
cannot be imbibed by reading 
or scholastic programmes 
but only through experiential 
learning and expert mentoring/
coaching.
attractive investment propositions. 
It is difficult (if not impossible) for 
incubators to engage more deeply 
with academics/researchers in labs 
and handhold them in crossing the 
early-stage valleys of death (e.g. 
finding proof of technology or 
proof of market). Incubators are 
vital for the ecosystem but their 
inbound supply chain needs to be 
strengthened.
3. Indian corporates and industries 
that are engaged with deep-tech 
startups do so only with those 
where technology is substantially 
developed or where revenues are visible. A majority 
of Indian corporates do not have knowledge or 
mechanisms for dealing with Open Innovation 
processes that our university/research institutions can 
potentially offer for creating deep-tech startups.
It is being proposed that policymakers should 
introduce Customer Discovery and Customer Development 
programmes to develop deep-tech startups from academic/
research institutions in India.
In 2013, the US Government through the National 
Science Foundation
4
 introduced the I-Corps programme
5
 
with great success to commercialise academic research in 
US universities. Quoting from NSF: “The I-Corps program 
uses experiential education to help researchers gain 
valuable insight into entrepreneurship, starting a business 
or industry requirements and challenges. I-Corps enables 
the transformation of invention to impact”. The most 
significant risk for startups is not failure of technology but 
failing to get adequate customers. The I-Corps programme 
is mandatory in the US for startups to obtain federal 
funding for research/commercialisation. 
Analogous to the I-Corps programme, the Government 
of India should consider making it mandatory for every 
translational research proposal at a university/research 
institution or a deep-tech startup seeking admission to a 
government incubator to undergo a rigorous Customer 
Discovery exercise. The learnings at such a programme 
can be truly transformative. 
The Gopalakrishnan-Deshpande Centre for Innovation 
& Entrepreneurship (GDC) at IIT Madras has successfully 
run its I-NCUBATE programme for the past four years 
and trained over 170 deep-tech startups from over 50 
colleges/incubators across India with excellent outcomes. 
The I-NCUBATE programme is inspired by the I-Corps 
programme. The empirical evidence of I-NCUBATE 
programme for success is described below:
1. Every participant startup in I-NCUBATE, without 
exception, found its innovation as not 
a good fit for the market. They would 
tweak their innovation or pivot it to 
become relevant. Two-thirds of startups 
found their early adopter customer 
segment in this manner. This puts 
them on a strong footing to build their 
prototype/MVP and provides insights 
into a good business model.
2. The remaining one-third of 
teams that do not find a “problem 
to solve” for their innovation have 
two outcomes post I-NCUBATE. 
Around 50% continue their Customer 
Discovery exercise and end up finding 
their early adopter customers. The residual 15% of 
teams conclude there is no problem to solve– i.e. their 
innovation is unlikely to succeed in the marketplace. 
This is not a failure (which is how incubators or 
investors would conclude) but actually a very good 
outcome for the researchers. Had they gone ahead 
with building their startup (without having done the 
I-NCUBATE programme), they would have spent 2-3 
years on it, spent money and other inputs and then 
encountered failure. 
3. The Customer Discovery exercise helps researchers 
know in 8 weeks (rather than learn it the hard way in 
3 years) if their innovation has a market, or how they 
should shape their startup journey to maximise chances 
for success. A “No-Go” is one of the best outcomes a 
researcher can get from the I-NCUBA TE programme. 
Conclusion 
Unfortunately, very few researchers and startup 
founders in India conduct a robust Customer Discovery 
exercise. This is more due to a lack of awareness and 
appreciation amongst policymakers of its transformational 
impact on the researchers/entrepreneurs. By linking 
development grants/seed investment programmes for 
deep-tech startups with a robust Customer Discovery 
exercise, we can create in India a significant amount of 
deal flow of robust and curated deep-tech startups into 
incubators and the ecosystem. More importantly, a fair 
share of deep-tech startups will help in solving India’s 
hard challenges.                                                              ?
(Views expressed in this article are personal.)
References
1. All data in this section (unless otherwise specified) is from 
NASSCOM Startup India Report – 2021.
2. CY – Calendar Year
3. Author’s estimate
4. National Science Foundation: https://www.nsf.gov/
5. https://www.nsf.gov/news/special_reports/i-corps/
Page 4


YOJANA   June 2022 17
ndia has a vibrant startup ecosystem with 
supporting infrastructure– incubators, 
development grants, angel/venture investors, 
mentors– and a conducive policy environment. 
The Economic Survey of India 2021-22 says that there 
are 61,400 registered startups in India, making it the  
third-largest startup ecosystem in the world behind China 
and US. Around 14,000
1
 new startups were registered 
in India during CY2021
2
. Over the past decade, Indian 
startups have created 6.6 lakh direct jobs and 34 lakh 
indirect jobs. 
Indian startups raised USD 24 billion in CY21, 
compared to USD 10 billion in CY20. There has been a 
significant localisation and diversification in the investor 
pool for startups in India over the past decade. There were 
more than 750 institutional investors in India in CY21, 
80% more than in CY20. The number of angel investors 
grew in CY21 by 20% to about 2,400. More than half the 
investment deals in CY21 had an India-based investor. 
Over 250 corporates have engaged with Indian startups 
in some way, including by running 80+ open innovation 
programmes for startups in CY21.
The Central and State governments in India have 
actively supported the startup sector over the past decade. 
The Startup India platform, which started in 2016, has 
been instrumental in encouraging startups and integrating 
them with the corporate and investment community. Over 
26 States in India have a startup policy.
What is a Deep-Tech Startup? 
Notwithstanding the healthy development of India’s 
startup ecosystem, one weakness that keeps India behind 
the developed countries is that we lack deep-tech startups. 
“Deep-tech” startups constitute less than one per cent of 
innovation Deep-Tech Startup Ecosystem
R Raghuttama Rao
The author is the CEO, Gopalakrishnan-Deshpande Centre for Innovation & Entrepreneurship, IIT Madras. Email: enquiries@gdciitm.org
I
the number of startups, far below what a fast-growing, 
complex, and large economy like India should have. 
The absence of deep-tech startups harms India 
considerably by weakening her capability to meaningfully 
address complex socio-economic challenges that afflict our 
society in multiple sectors such as agriculture, healthcare, 
transportation, education, energy, etc. The solutions to such 
challenges that address the UN’s Sustainable Development 
Goals would necessarily have to be radically new and 
disrupt existing industries and business processes.
In India’s population of 130 crores, only the top 
25%
3
 (affluent and middle-class) benefit from the fruits of 
technological progress, be it healthcare, consumer goods, 
clean water, safe transportation, education, etc. In contrast, 
the remaining 100 crore people do not get enough or are 
substantially bypassed. This is because most of the hi-tech 
goods and services are designed in the developed world 
for rich people– the average per capita income in OECD 
countries is about USD 40,000, while the average per 
capita income of the bottom 100 crore people in India is 
around USD 1000
3
. They simply cannot afford modern 
innovations with an income of 2.5% of the people for 
whom such innovations are designed. So, how do 100 
crore Indians move towards development?
The answer lies in becoming Atmanirbhar in 
commercialising domestic science and technology to solve 
our challenging problems.  
India’s development challenges are so unique and 
idiosyncratic that innovators from developed countries, 
not familiar with our context or cost structures, will not be 
able to provide solutions. The clarion call from the Prime 
Minister for ‘Atmanirbhar’ is apt here– we have to grow 
our own deep-tech ecosystem. 
Deep-tech startups arise from research-based, disruptive innovations from STEM labs of 
academic/research institutions and solve hard problems and challenges. India lacks deep-tech 
startups. Deep-tech startups constitute less than one per cent of the number of startups, far 
below what a fast-growing, complex, and large economy like India should have. 
18 YOJANA   June 2022
Need for Deep-Tech Startup Ecosystem
The phrase ‘deep-tech startup’ does not have a precise 
definition, but there is a broad consensus on what it is. 
Deep-tech startups arise from research-based, disruptive 
innovations from STEM labs of academic/research 
institutions and solve hard problems and challenges. 
Some examples are— (a) recycling sewage to get clean 
water at an affordable cost, (b) a low-cost solution at scale 
for curing blindness, (c) affordable solutions for treating 
diseases such as diabetes, dementia, cancer, etc., (d) 
creating an alternative to Lithium-ion batteries, and (e) 
low-cost satellite launching systems.
There are three major problems that deep-tech startups 
have vis-à-vis other startups (including those that are called 
tech-based startups).
1. Deep-tech startups need a longer gestation for 
development than other startups. The latter might 
need from 1-3 years to reach revenue, while deep-tech 
startups need 5-8 years.
2. Deep-tech startups require different types of inputs– 
they require more patient capital, specialised talent, 
and expert knowledge in more than one domain, to 
develop and validate a science-based innovation to the 
point where it is acceptable to commercial investors. 
For example, assume an invention involving creating 
a new substance (say a chemical that removes heavy 
metal from water). It takes time and resources to test 
and validate samples, obtain regulatory approvals, 
and set up a new manufacturing process to produce at 
scale. All these are capital-intensive, time-consuming, 
and have no assurance of success.
3. A deep-tech startup follows a different development 
path than other startups. A deep-tech startup derives 
its IP from the underlying science. The startup has to 
work backwards and find a real-life problem that is 
worth solving using its technology and validate the 
adequacy and nature of the market demand for the 
innovation.
Therefore, deep-tech startups 
take more time, talent, and capital 
to develop, upto when commercial 
investors find them acceptable. The 
risk of failure is high at every stage 
for a deep-tech startup, usually higher 
than in the case of other types of 
startups. But the payoffs of successful  
deep-tech startups are tremendous. 
Think of Microsoft, Google, Apple, 
Intel, Tesla, Moderna, SpaceX, etc. 
They are large corporations today, but 
they started as mere technology bets 
not very long ago. 
India has also created a few deep-tech startups over 
the past decade, whose impact has been overwhelmingly 
positive. It lends credence to the suggestion to step up 
policy and financial support to the deep-tech startup 
ecosystem.
Creating Ecosystem
India has produced about 94 unicorns so far, but barely 
any of them can claim to be a deep-tech startup. We have 
several venture funds in India, but most pursue relatively 
‘lower risk’ investment opportunities that exploit India’s 
growing consumption economy or those making cloned 
products. While India has a problem of inadequate R&D 
expenditure for an economy of her size, there is a sufficient 
amount of high-quality research in India’s top STEM 
colleges to fuel a deep-tech startup revolution. Some key 
reasons why our academic researchers lag in their potential 
to convert research into deep-tech startups are:
1. There is inadequate appreciation amongst 
policymakers and university administrators for the 
need to build capacity amongst academic researchers, 
scientists, and STEM students in India to truly 
understand what entrepreneurship entails and what 
commercialisation of research means. Being formally 
trained in science and technology but not having 
adequate exposure to the real world of business/
commerce, academic researchers conflate invention 
and innovation. There is a big difference between 
making a successful technological breakthrough in 
the lab and building a successful enterprise around 
it. Becoming entrepreneurial cannot be imbibed by 
reading or scholastic programmes but only through 
experiential learning and expert mentoring/coaching.
2. While Government has made good efforts to fund 
innovation in universities through programmes such 
as prototype development, filing for IPR, incubation, 
etc., few academics (<5%) commercialise their 
research by startups. A key point is that even if 
academics aspire to convert their inventions into 
enterprises, they do not have the mental make-up 
(the entrepreneur’s mindset) or the 
knowledge of how to organise what 
they have and collaborate with others to 
get what they do not have/know. Many 
universities have set up incubators 
to help with this, but they are not 
adequately equipped or incentivised to 
commercialise research. Although they 
are not-for-profit entities, incubators 
look for startups that have a good 
chance to be commercially viable. With 
their limited budgets, incubators face 
a tough challenge to nurture startups 
to scale their revenues and become 
Deep-tech startups take more 
time, talent, and capital 
to develop, upto when 
commercial investors find 
them acceptable. The risk of 
failure is high at every stage 
for a deep-tech startup, usually 
higher than in the case of 
other types of startups. But the 
payoffs of successful deep-tech 
startups are tremendous.
YOJANA   June 2022 19
There is a big difference 
between making a successful 
technological breakthrough 
in the lab and building a 
successful enterprise around 
it. Becoming entrepreneurial 
cannot be imbibed by reading 
or scholastic programmes 
but only through experiential 
learning and expert mentoring/
coaching.
attractive investment propositions. 
It is difficult (if not impossible) for 
incubators to engage more deeply 
with academics/researchers in labs 
and handhold them in crossing the 
early-stage valleys of death (e.g. 
finding proof of technology or 
proof of market). Incubators are 
vital for the ecosystem but their 
inbound supply chain needs to be 
strengthened.
3. Indian corporates and industries 
that are engaged with deep-tech 
startups do so only with those 
where technology is substantially 
developed or where revenues are visible. A majority 
of Indian corporates do not have knowledge or 
mechanisms for dealing with Open Innovation 
processes that our university/research institutions can 
potentially offer for creating deep-tech startups.
It is being proposed that policymakers should 
introduce Customer Discovery and Customer Development 
programmes to develop deep-tech startups from academic/
research institutions in India.
In 2013, the US Government through the National 
Science Foundation
4
 introduced the I-Corps programme
5
 
with great success to commercialise academic research in 
US universities. Quoting from NSF: “The I-Corps program 
uses experiential education to help researchers gain 
valuable insight into entrepreneurship, starting a business 
or industry requirements and challenges. I-Corps enables 
the transformation of invention to impact”. The most 
significant risk for startups is not failure of technology but 
failing to get adequate customers. The I-Corps programme 
is mandatory in the US for startups to obtain federal 
funding for research/commercialisation. 
Analogous to the I-Corps programme, the Government 
of India should consider making it mandatory for every 
translational research proposal at a university/research 
institution or a deep-tech startup seeking admission to a 
government incubator to undergo a rigorous Customer 
Discovery exercise. The learnings at such a programme 
can be truly transformative. 
The Gopalakrishnan-Deshpande Centre for Innovation 
& Entrepreneurship (GDC) at IIT Madras has successfully 
run its I-NCUBATE programme for the past four years 
and trained over 170 deep-tech startups from over 50 
colleges/incubators across India with excellent outcomes. 
The I-NCUBATE programme is inspired by the I-Corps 
programme. The empirical evidence of I-NCUBATE 
programme for success is described below:
1. Every participant startup in I-NCUBATE, without 
exception, found its innovation as not 
a good fit for the market. They would 
tweak their innovation or pivot it to 
become relevant. Two-thirds of startups 
found their early adopter customer 
segment in this manner. This puts 
them on a strong footing to build their 
prototype/MVP and provides insights 
into a good business model.
2. The remaining one-third of 
teams that do not find a “problem 
to solve” for their innovation have 
two outcomes post I-NCUBATE. 
Around 50% continue their Customer 
Discovery exercise and end up finding 
their early adopter customers. The residual 15% of 
teams conclude there is no problem to solve– i.e. their 
innovation is unlikely to succeed in the marketplace. 
This is not a failure (which is how incubators or 
investors would conclude) but actually a very good 
outcome for the researchers. Had they gone ahead 
with building their startup (without having done the 
I-NCUBATE programme), they would have spent 2-3 
years on it, spent money and other inputs and then 
encountered failure. 
3. The Customer Discovery exercise helps researchers 
know in 8 weeks (rather than learn it the hard way in 
3 years) if their innovation has a market, or how they 
should shape their startup journey to maximise chances 
for success. A “No-Go” is one of the best outcomes a 
researcher can get from the I-NCUBA TE programme. 
Conclusion 
Unfortunately, very few researchers and startup 
founders in India conduct a robust Customer Discovery 
exercise. This is more due to a lack of awareness and 
appreciation amongst policymakers of its transformational 
impact on the researchers/entrepreneurs. By linking 
development grants/seed investment programmes for 
deep-tech startups with a robust Customer Discovery 
exercise, we can create in India a significant amount of 
deal flow of robust and curated deep-tech startups into 
incubators and the ecosystem. More importantly, a fair 
share of deep-tech startups will help in solving India’s 
hard challenges.                                                              ?
(Views expressed in this article are personal.)
References
1. All data in this section (unless otherwise specified) is from 
NASSCOM Startup India Report – 2021.
2. CY – Calendar Year
3. Author’s estimate
4. National Science Foundation: https://www.nsf.gov/
5. https://www.nsf.gov/news/special_reports/i-corps/
YOJANA   June 2022 23
ndia was ranked 8
th
 in the top 10 countries 
by AI patent families
1
, ahead of Russia and 
France, with AI-related patent applications 
growing tenfold from 2012 to 2018. A report 
titled ‘AI Enabled SaaS: The Next Frontier for Global 
SaaS Start-ups from India’ highlighted that AI could 
generate over 9,00,000 white-collar jobs and 3.6 million 
indirect jobs by 2030. Additionally, India has over 1,300 
Global Capability Centres (GCCs), with one in every 
five GCCs using AI across key business functions such 
as cybersecurity, customer services, supply chain, and 
operations management.
2
To support this thriving AI industry, India is one of 
the handful of countries that have developed a conceptual 
national framework for the use of artificial intelligence 
(AI) and its allied field, machine learning. The Government 
expenditure on AI and Machine Learning has also steadily 
increased with expected growth at a CAGR of 39 per cent 
over the period 2019-2025 to reach USD 11,781.9 million 
in 2025
3
. As per the government’s think tank NITI Aayog, 
by 2035, AI has the potential to add USD 1 trillion to the 
Indian economy. However, successful adoption of AI will 
require strategy, implementation, risk management, and an 
AI-enabled workforce. Even to promote innovations in AI, 
a streamlined national policy framework is necessary.
In this regard, by recognising AI’s potential to 
transform the Indian economy and the need for India to 
AI and Machine Learning
Hindol Sengupta 
Bhavya Tyagi 
Hindol Sengupta is V ice President and Head of the Strategic Investment Research Unit at Invest India. Email: hindol.sengupta@investindia.org.in
Bhavya Tyagi is a researcher at the same organisation.
Industry 4.0 is set to usher an era of technologies that will completely alter the way in which we 
interact with the world around us. Artificial Intelligence/Machine Learning, IoT, 5G, Augmented 
Reality, Big Data, Nanotechnology, Robotics, and 3D printing are transforming the operational, 
functional, and strategic landscape across various industries. In India, both private and public 
enterprises and the Central and State governments are investing in multiple AI use cases– from 
manufacturing to services. Most venture capital funding in India is now going to AI projects in 
Banking, Financial Services, and Insurance Sector (BFSI), e-commerce, healthcare, electronics 
and renewable energy startups.
applications I
build a comprehensive strategic framework to harness 
it, NITI Aayog released a National Strategy for Artificial 
Intelligence #AIforAll in June 2018. The paper lays 
out the roadmap for India to leverage the coming-of-
age technologies to ensure inclusive growth and social 
development. #AIforAll aims at enhancing and empowering 
human capabilities to address challenges of access, 
affordability, and efficiency in endeavouring to scale Made-
fOCuS
Page 5


YOJANA   June 2022 17
ndia has a vibrant startup ecosystem with 
supporting infrastructure– incubators, 
development grants, angel/venture investors, 
mentors– and a conducive policy environment. 
The Economic Survey of India 2021-22 says that there 
are 61,400 registered startups in India, making it the  
third-largest startup ecosystem in the world behind China 
and US. Around 14,000
1
 new startups were registered 
in India during CY2021
2
. Over the past decade, Indian 
startups have created 6.6 lakh direct jobs and 34 lakh 
indirect jobs. 
Indian startups raised USD 24 billion in CY21, 
compared to USD 10 billion in CY20. There has been a 
significant localisation and diversification in the investor 
pool for startups in India over the past decade. There were 
more than 750 institutional investors in India in CY21, 
80% more than in CY20. The number of angel investors 
grew in CY21 by 20% to about 2,400. More than half the 
investment deals in CY21 had an India-based investor. 
Over 250 corporates have engaged with Indian startups 
in some way, including by running 80+ open innovation 
programmes for startups in CY21.
The Central and State governments in India have 
actively supported the startup sector over the past decade. 
The Startup India platform, which started in 2016, has 
been instrumental in encouraging startups and integrating 
them with the corporate and investment community. Over 
26 States in India have a startup policy.
What is a Deep-Tech Startup? 
Notwithstanding the healthy development of India’s 
startup ecosystem, one weakness that keeps India behind 
the developed countries is that we lack deep-tech startups. 
“Deep-tech” startups constitute less than one per cent of 
innovation Deep-Tech Startup Ecosystem
R Raghuttama Rao
The author is the CEO, Gopalakrishnan-Deshpande Centre for Innovation & Entrepreneurship, IIT Madras. Email: enquiries@gdciitm.org
I
the number of startups, far below what a fast-growing, 
complex, and large economy like India should have. 
The absence of deep-tech startups harms India 
considerably by weakening her capability to meaningfully 
address complex socio-economic challenges that afflict our 
society in multiple sectors such as agriculture, healthcare, 
transportation, education, energy, etc. The solutions to such 
challenges that address the UN’s Sustainable Development 
Goals would necessarily have to be radically new and 
disrupt existing industries and business processes.
In India’s population of 130 crores, only the top 
25%
3
 (affluent and middle-class) benefit from the fruits of 
technological progress, be it healthcare, consumer goods, 
clean water, safe transportation, education, etc. In contrast, 
the remaining 100 crore people do not get enough or are 
substantially bypassed. This is because most of the hi-tech 
goods and services are designed in the developed world 
for rich people– the average per capita income in OECD 
countries is about USD 40,000, while the average per 
capita income of the bottom 100 crore people in India is 
around USD 1000
3
. They simply cannot afford modern 
innovations with an income of 2.5% of the people for 
whom such innovations are designed. So, how do 100 
crore Indians move towards development?
The answer lies in becoming Atmanirbhar in 
commercialising domestic science and technology to solve 
our challenging problems.  
India’s development challenges are so unique and 
idiosyncratic that innovators from developed countries, 
not familiar with our context or cost structures, will not be 
able to provide solutions. The clarion call from the Prime 
Minister for ‘Atmanirbhar’ is apt here– we have to grow 
our own deep-tech ecosystem. 
Deep-tech startups arise from research-based, disruptive innovations from STEM labs of 
academic/research institutions and solve hard problems and challenges. India lacks deep-tech 
startups. Deep-tech startups constitute less than one per cent of the number of startups, far 
below what a fast-growing, complex, and large economy like India should have. 
18 YOJANA   June 2022
Need for Deep-Tech Startup Ecosystem
The phrase ‘deep-tech startup’ does not have a precise 
definition, but there is a broad consensus on what it is. 
Deep-tech startups arise from research-based, disruptive 
innovations from STEM labs of academic/research 
institutions and solve hard problems and challenges. 
Some examples are— (a) recycling sewage to get clean 
water at an affordable cost, (b) a low-cost solution at scale 
for curing blindness, (c) affordable solutions for treating 
diseases such as diabetes, dementia, cancer, etc., (d) 
creating an alternative to Lithium-ion batteries, and (e) 
low-cost satellite launching systems.
There are three major problems that deep-tech startups 
have vis-à-vis other startups (including those that are called 
tech-based startups).
1. Deep-tech startups need a longer gestation for 
development than other startups. The latter might 
need from 1-3 years to reach revenue, while deep-tech 
startups need 5-8 years.
2. Deep-tech startups require different types of inputs– 
they require more patient capital, specialised talent, 
and expert knowledge in more than one domain, to 
develop and validate a science-based innovation to the 
point where it is acceptable to commercial investors. 
For example, assume an invention involving creating 
a new substance (say a chemical that removes heavy 
metal from water). It takes time and resources to test 
and validate samples, obtain regulatory approvals, 
and set up a new manufacturing process to produce at 
scale. All these are capital-intensive, time-consuming, 
and have no assurance of success.
3. A deep-tech startup follows a different development 
path than other startups. A deep-tech startup derives 
its IP from the underlying science. The startup has to 
work backwards and find a real-life problem that is 
worth solving using its technology and validate the 
adequacy and nature of the market demand for the 
innovation.
Therefore, deep-tech startups 
take more time, talent, and capital 
to develop, upto when commercial 
investors find them acceptable. The 
risk of failure is high at every stage 
for a deep-tech startup, usually higher 
than in the case of other types of 
startups. But the payoffs of successful  
deep-tech startups are tremendous. 
Think of Microsoft, Google, Apple, 
Intel, Tesla, Moderna, SpaceX, etc. 
They are large corporations today, but 
they started as mere technology bets 
not very long ago. 
India has also created a few deep-tech startups over 
the past decade, whose impact has been overwhelmingly 
positive. It lends credence to the suggestion to step up 
policy and financial support to the deep-tech startup 
ecosystem.
Creating Ecosystem
India has produced about 94 unicorns so far, but barely 
any of them can claim to be a deep-tech startup. We have 
several venture funds in India, but most pursue relatively 
‘lower risk’ investment opportunities that exploit India’s 
growing consumption economy or those making cloned 
products. While India has a problem of inadequate R&D 
expenditure for an economy of her size, there is a sufficient 
amount of high-quality research in India’s top STEM 
colleges to fuel a deep-tech startup revolution. Some key 
reasons why our academic researchers lag in their potential 
to convert research into deep-tech startups are:
1. There is inadequate appreciation amongst 
policymakers and university administrators for the 
need to build capacity amongst academic researchers, 
scientists, and STEM students in India to truly 
understand what entrepreneurship entails and what 
commercialisation of research means. Being formally 
trained in science and technology but not having 
adequate exposure to the real world of business/
commerce, academic researchers conflate invention 
and innovation. There is a big difference between 
making a successful technological breakthrough in 
the lab and building a successful enterprise around 
it. Becoming entrepreneurial cannot be imbibed by 
reading or scholastic programmes but only through 
experiential learning and expert mentoring/coaching.
2. While Government has made good efforts to fund 
innovation in universities through programmes such 
as prototype development, filing for IPR, incubation, 
etc., few academics (<5%) commercialise their 
research by startups. A key point is that even if 
academics aspire to convert their inventions into 
enterprises, they do not have the mental make-up 
(the entrepreneur’s mindset) or the 
knowledge of how to organise what 
they have and collaborate with others to 
get what they do not have/know. Many 
universities have set up incubators 
to help with this, but they are not 
adequately equipped or incentivised to 
commercialise research. Although they 
are not-for-profit entities, incubators 
look for startups that have a good 
chance to be commercially viable. With 
their limited budgets, incubators face 
a tough challenge to nurture startups 
to scale their revenues and become 
Deep-tech startups take more 
time, talent, and capital 
to develop, upto when 
commercial investors find 
them acceptable. The risk of 
failure is high at every stage 
for a deep-tech startup, usually 
higher than in the case of 
other types of startups. But the 
payoffs of successful deep-tech 
startups are tremendous.
YOJANA   June 2022 19
There is a big difference 
between making a successful 
technological breakthrough 
in the lab and building a 
successful enterprise around 
it. Becoming entrepreneurial 
cannot be imbibed by reading 
or scholastic programmes 
but only through experiential 
learning and expert mentoring/
coaching.
attractive investment propositions. 
It is difficult (if not impossible) for 
incubators to engage more deeply 
with academics/researchers in labs 
and handhold them in crossing the 
early-stage valleys of death (e.g. 
finding proof of technology or 
proof of market). Incubators are 
vital for the ecosystem but their 
inbound supply chain needs to be 
strengthened.
3. Indian corporates and industries 
that are engaged with deep-tech 
startups do so only with those 
where technology is substantially 
developed or where revenues are visible. A majority 
of Indian corporates do not have knowledge or 
mechanisms for dealing with Open Innovation 
processes that our university/research institutions can 
potentially offer for creating deep-tech startups.
It is being proposed that policymakers should 
introduce Customer Discovery and Customer Development 
programmes to develop deep-tech startups from academic/
research institutions in India.
In 2013, the US Government through the National 
Science Foundation
4
 introduced the I-Corps programme
5
 
with great success to commercialise academic research in 
US universities. Quoting from NSF: “The I-Corps program 
uses experiential education to help researchers gain 
valuable insight into entrepreneurship, starting a business 
or industry requirements and challenges. I-Corps enables 
the transformation of invention to impact”. The most 
significant risk for startups is not failure of technology but 
failing to get adequate customers. The I-Corps programme 
is mandatory in the US for startups to obtain federal 
funding for research/commercialisation. 
Analogous to the I-Corps programme, the Government 
of India should consider making it mandatory for every 
translational research proposal at a university/research 
institution or a deep-tech startup seeking admission to a 
government incubator to undergo a rigorous Customer 
Discovery exercise. The learnings at such a programme 
can be truly transformative. 
The Gopalakrishnan-Deshpande Centre for Innovation 
& Entrepreneurship (GDC) at IIT Madras has successfully 
run its I-NCUBATE programme for the past four years 
and trained over 170 deep-tech startups from over 50 
colleges/incubators across India with excellent outcomes. 
The I-NCUBATE programme is inspired by the I-Corps 
programme. The empirical evidence of I-NCUBATE 
programme for success is described below:
1. Every participant startup in I-NCUBATE, without 
exception, found its innovation as not 
a good fit for the market. They would 
tweak their innovation or pivot it to 
become relevant. Two-thirds of startups 
found their early adopter customer 
segment in this manner. This puts 
them on a strong footing to build their 
prototype/MVP and provides insights 
into a good business model.
2. The remaining one-third of 
teams that do not find a “problem 
to solve” for their innovation have 
two outcomes post I-NCUBATE. 
Around 50% continue their Customer 
Discovery exercise and end up finding 
their early adopter customers. The residual 15% of 
teams conclude there is no problem to solve– i.e. their 
innovation is unlikely to succeed in the marketplace. 
This is not a failure (which is how incubators or 
investors would conclude) but actually a very good 
outcome for the researchers. Had they gone ahead 
with building their startup (without having done the 
I-NCUBATE programme), they would have spent 2-3 
years on it, spent money and other inputs and then 
encountered failure. 
3. The Customer Discovery exercise helps researchers 
know in 8 weeks (rather than learn it the hard way in 
3 years) if their innovation has a market, or how they 
should shape their startup journey to maximise chances 
for success. A “No-Go” is one of the best outcomes a 
researcher can get from the I-NCUBA TE programme. 
Conclusion 
Unfortunately, very few researchers and startup 
founders in India conduct a robust Customer Discovery 
exercise. This is more due to a lack of awareness and 
appreciation amongst policymakers of its transformational 
impact on the researchers/entrepreneurs. By linking 
development grants/seed investment programmes for 
deep-tech startups with a robust Customer Discovery 
exercise, we can create in India a significant amount of 
deal flow of robust and curated deep-tech startups into 
incubators and the ecosystem. More importantly, a fair 
share of deep-tech startups will help in solving India’s 
hard challenges.                                                              ?
(Views expressed in this article are personal.)
References
1. All data in this section (unless otherwise specified) is from 
NASSCOM Startup India Report – 2021.
2. CY – Calendar Year
3. Author’s estimate
4. National Science Foundation: https://www.nsf.gov/
5. https://www.nsf.gov/news/special_reports/i-corps/
YOJANA   June 2022 23
ndia was ranked 8
th
 in the top 10 countries 
by AI patent families
1
, ahead of Russia and 
France, with AI-related patent applications 
growing tenfold from 2012 to 2018. A report 
titled ‘AI Enabled SaaS: The Next Frontier for Global 
SaaS Start-ups from India’ highlighted that AI could 
generate over 9,00,000 white-collar jobs and 3.6 million 
indirect jobs by 2030. Additionally, India has over 1,300 
Global Capability Centres (GCCs), with one in every 
five GCCs using AI across key business functions such 
as cybersecurity, customer services, supply chain, and 
operations management.
2
To support this thriving AI industry, India is one of 
the handful of countries that have developed a conceptual 
national framework for the use of artificial intelligence 
(AI) and its allied field, machine learning. The Government 
expenditure on AI and Machine Learning has also steadily 
increased with expected growth at a CAGR of 39 per cent 
over the period 2019-2025 to reach USD 11,781.9 million 
in 2025
3
. As per the government’s think tank NITI Aayog, 
by 2035, AI has the potential to add USD 1 trillion to the 
Indian economy. However, successful adoption of AI will 
require strategy, implementation, risk management, and an 
AI-enabled workforce. Even to promote innovations in AI, 
a streamlined national policy framework is necessary.
In this regard, by recognising AI’s potential to 
transform the Indian economy and the need for India to 
AI and Machine Learning
Hindol Sengupta 
Bhavya Tyagi 
Hindol Sengupta is V ice President and Head of the Strategic Investment Research Unit at Invest India. Email: hindol.sengupta@investindia.org.in
Bhavya Tyagi is a researcher at the same organisation.
Industry 4.0 is set to usher an era of technologies that will completely alter the way in which we 
interact with the world around us. Artificial Intelligence/Machine Learning, IoT, 5G, Augmented 
Reality, Big Data, Nanotechnology, Robotics, and 3D printing are transforming the operational, 
functional, and strategic landscape across various industries. In India, both private and public 
enterprises and the Central and State governments are investing in multiple AI use cases– from 
manufacturing to services. Most venture capital funding in India is now going to AI projects in 
Banking, Financial Services, and Insurance Sector (BFSI), e-commerce, healthcare, electronics 
and renewable energy startups.
applications I
build a comprehensive strategic framework to harness 
it, NITI Aayog released a National Strategy for Artificial 
Intelligence #AIforAll in June 2018. The paper lays 
out the roadmap for India to leverage the coming-of-
age technologies to ensure inclusive growth and social 
development. #AIforAll aims at enhancing and empowering 
human capabilities to address challenges of access, 
affordability, and efficiency in endeavouring to scale Made-
fOCuS
24 YOJANA   June 2022
AI ecosystem in the government 
currently comprises capacity 
building and reskilling, policies, 
innovation centres, and projects. 
However, given the agile 
nature of the technologies, 
several research reports have 
pressed for the development 
and deployment of an agile 
regulatory and infrastructural 
framework to remove 
bottlenecks and complexities in 
AI-driven processes and ensure 
standardisation of AI. 
in-India artificial intelligence solutions 
for the benefit of the developing and 
emerging economies. Here, ‘solve 
for India’ means to ‘solve for 40 per 
cent of the world.’ In doing so, the 
paper identifies five priority sectors 
that are envisioned to gain the most 
incremental value from the adoption 
of these transformative technologies in 
solving societal needs: a) healthcare– 
increasing access to quality and 
affordable healthcare, b) agriculture– 
enhancing crop yield, ensuring food 
security, and increasing farmers’ 
incomes, c) education– enhancing 
the quality of education and human 
resource productivity, d) smart cities 
and infrastructure– ensuring efficient 
connectivity and promoting intelligent urban planning, and 
e) smart mobility and transportation– enabling efficient and 
safe transportation. The National Strategy aims to support 
and enable India’s AI ecosystem through grants, product 
and solution development, collaboration with the industry, 
and mentorship support to startups. Since the release of 
this report, NITI Aayog has launched several initiatives 
such as the Atal Innovation Mission, Empowered Group-6, 
RAISE 2020 summit, etc. The flagship AI initiative of NITI 
Aayog has been the Responsible AI Approach Documents 
published in collaboration with the World Economic 
Forum Centre for the next AI Industrial Revolution. The 
Documents seek to establish broad ethics and principles for 
the design, development, and deployment of AI in India. 
AI ecosystem in the government currently comprises 
capacity building and reskilling, policies, innovation 
centres, and projects. However, given the agile nature of 
the technologies, several research reports have pressed for 
the development and deployment of an agile regulatory 
and infrastructural framework to remove bottlenecks 
and complexities in AI-driven processes and ensure 
standardization of AI. Keeping this in mind, India’s 
AI Standardization Committee of the Department of 
Telecommunications (DoT) released a draft framework on 
the India Artificial Intelligence Stack 
to enable an environment to exploit 
AI productively across all sectors and 
bring interoperability, among other 
things. The Stack is divided into six 
layers– five main horizontal layers 
and one vertical layer— each catering 
to a specific purpose ranging from 
information gathering and storage to 
security and governance. The Report 
highlights numerous benefits of the AI 
Stack such as secure data storage and 
data privacy, easy interface, protection 
of digital rights, open API integration, 
trustworthiness, ethical standards, 
and usage of government Public Key 
Infrastructure (PKI). It further lays 
down the roadmap to increase public 
and private partnership in research, 
accelerating adoption of AI, skilling 
the workforce, and ensuring ethics 
and principle for a responsible AI. It 
also ensures the creation of a common 
Data controller, including multi-cloud 
scenarios. One of the major advantages 
of this proposed AI Stack is that it will 
facilitate open API integration and 
build AI architecture from square one. 
Through this, the Government aims to 
provide a balanced ‘playground’ for 
institutions to accelerate research and development in AI 
and ensure speedy adoption of AI across the value chain.
India is bringing in the use of AI in everything– from 
promoting digital health, and amplifying digital financial 
transactions to helping pensioners receive their payments 
with greater ease, and tracking down tigers to preserve 
them. 
In Telangana, AI is helping authenticate pensioners 
and ensuring that payments go to pensioners who are alive 
(thus, removing chances of graft) and using basic images 
and information to help validate recipients.
The Ministry of Corporate Affairs is using AI to 
simplify corporate filings, while the Centre for Artificial 
Intelligence and Robotics (CAIR) has been built as a 
special hub for AI-related work of the DRDO (Defence 
Research and Development Organisation). AI is an area of 
special importance for the National Research Foundation, 
and it is being promoted at the school level to encourage 
new talent in this sphere. 
India sees AI and machine learning as the next 
transformative process to reform its economy and give it 
greater depth, and weed out irregularities. This process has 
already started with the digitisation of citizen identity and 
financial transactions, but the use of AI 
would give it a whole new dimension 
and depth. 
AI and machine learning are 
particularly suited for India because the 
country is the world’s largest generator 
of democratic data or data which is 
being generated and analysed under the 
democratic rule of governance. 
This wealth of data gives India 
an advantage in many fields– from 
India is bringing in the use 
of AI in everything from 
promoting digital health, and 
amplifying digital financial 
transactions to helping 
pensioners receive their 
payments with greater ease, 
and tracking down tigers to 
preserve them.